The electrocatalytic hydrogenation of furanic compounds in a continuous electrocatalytic membrane reactor

Article abstract of DOI:10.1039/c3gc00090g

The electrocatalytic hydrogenation of biomass derived oxygenates in a continuous electrocatalytic membrane reactor presents a promising method of fuel and chemical production that minimizes usage of solvents and has the potential to be powered using renewable electricity. In this paper we demonstrate the use of a continuous-flow electrocatalytic membrane reactor for the reduction of aqueous solutions of furfural into furfuryl alcohol (FA), tetrahydrofurfuryl alcohol (THFA), 2-methylfuran (MF) and 2-methyltetrahydrofuran (MTHF). Protons needed for hydrogenation were obtained from the electrolysis of water at the anode of the reactor. Pd was identified as the most active monometallic catalyst of 5 different catalysts tested for the hydrogenation of aqueous furfural with hydrogen gas in a high-throughput reactor. Thus Pd/C was tested as a cathode catalyst for the electrocatalytic hydrogenation of furfural. At a power input of 0.1W, Pd/C was 4.4 times more active (per active metal site) as a cathode catalyst in the electrocatalytic hydrogenation of furfural than Pt/C. The main products for the electrocatalytic hydrogenation of furfural were FA (54-100% selectivity) and THFA (0-26% selectivity). MF and MTHF were also detected in selectivities of 8%. Varying the reactor temperature between 30 °C and 70 °C had a minimal effect on reaction rate for furfural conversion. Using hydrogen gas at the anode, in place of water electrolysis, produced slightly higher rates of product formation at a lower power input. Sparging hydrogen gas on the cathode had no effect on reaction rate or selectivity, and was used to examine the addition of recycling loops to the continuous electrocatalytic membrane reactor.

Full text of DOI:10.1039/c3gc00090g

Green Chemistry
View Article Online
View Journal
PAPER
The electrocatalytic hydrogenation of furanic
compounds in a continuous electrocatalytic membrane
Cite this: DOI: 10.1039/c3gc00090g
reactor†
a
a
a
a
Sara K. Green, Jechan Lee,‡ Hyung Ju Kim,‡ Geoffrey A. Tompsett,
b
a
Won Bae Kim and George W. Huber‡*
The electrocatalytic hydrogenation of biomass derived oxygenates in a continuous electrocatalytic mem-
brane reactor presents a promising method of fuel and chemical production that minimizes usage of sol-
vents and has the potential to be powered using renewable electricity. In this paper we demonstrate the
use of a continuous-flow electrocatalytic membrane reactor for the reduction of aqueous solutions of
furfural into furfuryl alcohol (FA), tetrahydrofurfuryl alcohol (THFA), 2-methylfuran (MF) and 2-methyl-
tetrahydrofuran (MTHF). Protons needed for hydrogenation were obtained from the electrolysis of water
at the anode of the reactor. Pd was identified as the most active monometallic catalyst of 5 different cata-
lysts tested for the hydrogenation of aqueous furfural with hydrogen gas in a high-throughput reactor.
Thus Pd/C was tested as a cathode catalyst for the electrocatalytic hydrogenation of furfural. At a power
input of 0.1W, Pd/C was 4.4 times more active (per active metal site) as a cathode catalyst in the electro-
catalytic hydrogenation of furfural than Pt/C. The main products for the electrocatalytic hydrogenation
of furfural were FA (54–100% selectivity) and THFA (0–26% selectivity). MF and MTHF were also detected
in selectivities of 8%. Varying the reactor temperature between 30 °C and 70 °C had a minimal effect on
reaction rate for furfural conversion. Using hydrogen gas at the anode, in place of water electrolysis, pro-
duced slightly higher rates of product formation at a lower power input. Sparging hydrogen gas on the
cathode had no effect on reaction rate or selectivity, and was used to examine the addition of recycling
loops to the continuous electrocatalytic membrane reactor.
Received 12th January 2013,
Accepted 9th May 2013
DOI: 10.1039/c3gc00090g
www.rsc.org/greenchem
as $0.13 per gallon diesel fuel for hydrodeoxygenation (HDO)
of vegetable oils and $1.00 per gallon of gasoline for the HDO
Introduction
5
–8
Many current methods of biomass conversion require the use of bio-oils and sugars.
These costs amount to a significant
of high quantities of hydrogen gas. This usage of hydrogen gas portion of the overall price of the product. This example illus-
constitutes a large fraction of the overall operating cost of trates the need for methods of biomass conversion that mini-
1
–4
these technologies.
An example is the hydrodeoxygenation mize consumption of hydrogen gas.
We recently demonstrated that proton exchange membrane
per kg feedstock. The cost of hydro- (PEM) technology can be used to electrocatalytically hydrogen-
gen gas is about $1 per kg near hydrogen pipelines, but can ate acetone to isopropyl alcohol (IPA) using protons generated
of carbohydrates, vegetable oils, and pyrolysis oils, which can
5
–8
require up to 0.12 kg H
2
1
0
increase 10 times or more when storage and transportation are from the electrolysis of water. Water electrolysis requires an
9
required. This is especially important in remote locations, electrical power input (defined as the product of applied
where biomass is plentiful and inexpensive. Assuming $2 per voltage and measured current). The required cell potential for
kg H , the cost of the hydrogen gas feedstock could be as high water electrolysis is proportional to the Gibbs free energy via
2
the number of electrons transferred and Faraday’s constant
(
ΔG = −nFECell). Cell current has a complicated relationship
a
with cell potential, which was simplified by Datta into a model
Department of Chemical Engineering, University of Massachusetts-Amherst, 686
based on charge balances, mass balances, and Butler–Volmer
kinetics.
The electrocatalytic hydrogenation of biomass-derived oxy-
genates has several key advantages over other methods of
biomass conversion. The most significant advantage is that no
hydrogen gas is required. This is especially important in
North Pleasant Street, Amherst, MA 01003, USA. E-mail: huber@engr.wisc.edu
b
11
School of Materials Science and Engineering, Gwangju Institute of Science and
Technology (GIST), 261 Cheomdan-gwagiro, Buk-gu, Gwangju, 500-712, South Korea
†
1
‡
Electronic supplementary information (ESI)
0.1039/c3gc00090g
available. See DOI:
Current address: Department of Chemical and Biological Engineering, Univer-
sity of Wisconsin-Madison, 1415 Engineering Drive, Madison, WI 53706, USA.
This journal is © The Royal Society of Chemistry 2013
Green Chem.

View Article Online
Paper
Green Chemistry
9
remote locations where hydrogen gas is expensive. Another (THFA), 2-methylfuran (MF), and 2-methyltetrahydrofuran
advantage is that this technology stores electricity as a liquid (MTHF). Furfural is produced from the acid-catalyzed de-
1
2,13
transportation fuel that fits into existing infrastructure.
hydration of xylose at around 300 000 metric tons per year, and
This technology could thus be used for electricity storage, cannot be used directly as a fuel due to its tendency to poly-
2
1
especially when the electricity is generated from renewable merize at room temperatures. However, MTHF has an octane
1
4
sources like wind turbines. Compared to the fermentation of number of 74 and can be blended into p-series gasoline
1
,22
glucose to ethanol, the electrocatalytic HDO of glucose to fuel.
Additionally, THFA is a “green” solvent used in indus-
hexane in a continuous electrocatalytic membrane reactor has trial applications such as biocides and pesticides, coatings
1
0
25
the potential to produce 50% more energy as a liquid fuel.
and dyes, and stripping formulations. The Gibbs’s free ener-
HDO of biomass produces a wide range of products including gies and standard cell potentials for the electrocatalytic for-
alcohols, polyols, and alkanes that can be used for fuel blend- mation of these products from furfural are presented in
5
,15,16
stocks or sold into the commodity chemicals market.
Table 1. As mentioned above, reactions using protons gener-
Thus PEM technology could be used to produce a wide range ated from water electrolysis require a higher cell potential, and
of products from the electrocatalytic hydrogenation of thus a greater power input. The formation of THFA and MF
biomass-derived molecules. Both the anodic water electrolysis from FA requires 1.01 V and 1.09 V, respectively. This indicates
and the cathodic biomass hydrogenation take place in one that hydrogenating the furan ring to form THFA is marginally
single continuous reactor that could easily be scaled up to a more favorable than hydrogenolysis of the conjugated hydroxyl
commercial level. Having one single reactor vessel could group to form MF. Similarly, hydrogenating the furan ring in
reduce the capital cost of a facility for biomass conversion. MF to form MTHF (1.5 V) is more favorable than hydrogeno-
The simplicity of the reactor design would also decrease the lysis of the non-conjugated hydroxyl group in THFA to form
operating costs. Finally, one of the most important advantages MTHF (2.06 V). In fact, though MTHF can be produced from
of electrocatalysis compared to conventional catalysis is the THFA in conventional catalysis, this route is highly unlikely to
ability to manipulate the size of the activation barrier by con- occur electrocatalytically because the formation of hydrogen
1
7,18
trolling the electrode potential.
Therefore, reactions that gas, requiring 1.23 V, is significantly more thermodynamically
require high temperatures and pressures in a conventional cat- favorable and, subsequently, faster.
alytic system (e.g., hydrolysis of water) will readily occur at
Several catalysts have been studied as replacements for the
atmospheric temperatures and pressures in the continuous toxic copper chromate catalyst historically used for furfural
electrocatalytic membrane reactor due to the application of a hydrogenation. These catalysts include: carbon supported
1
9,20
26,27
28
29
voltage.
Additionally, controlling the electrode potential copper,
a copper/iron catalyst, RANEY® Ni, and mono-
2
5,30–33
has the ability to affect reaction selectivity, as high potentials metallic and bimetallic Pt, Pd, Sn, and Ru.
These cata-
may promote the formation of secondary or side-products. lysts can also be employed in electrocatalysis. Miller, Jackson,
Based on the required power input, electrocatalytic hydrogen- and Saffron explored furfural hydrogenation to FA and MF in
ation in a continuous electrocatalytic membrane reactor will an undivided electrochemical cell using a sacrificial Ni or Ni–
be most economically profitable when hydrogen gas is expen- Fe alloy anode with an Al, Fe, Ni, or Cu cathode, and an
3
4
sive and electricity is cheap.
ammonium chloride electrolyte. FA production was favored
Furfural is a, commercially relevant, platform molecule that at pH of 5.0 while MF production was favored at pH 1.0.
can produce a wide range of products by hydrogenation reac- Increasing the current density decreased the reaction rate and
tions (Fig. 1) and shows significant potential for electrocataly- efficiency. Chu et al. demonstrated the electrocatalytic hydro-
1
,21–24
tic hydrogenation.
2
Products from furfural hydrogenation genation of furfural using a nanoporous TiO catalyst with
3
5
include furfuryl alcohol (FA), tetrahydrofurfuryl alcohol enhanced surface area in an ionic liquid. Belgsir has investi-
gated the electrocatalytic conversion of furfural using both a
3
6,37
flow-through reactor and a half-cell batch reactor.
In the
flow-through reactor, the oxidation of furfural to furonic acid
over a nickel modified graphite felt electrode was coupled with
Table 1 Gibbs free energy and standard cell potentials for the hydrogenation
of furfural using hydrogen gas and water electrolysis
Hydrogen gas
Water electrolysis
cell (V) ΔG (kJ mol−1
Ecell (V)
1
Reaction
ΔG (kJ mol−
)
E
)
1
2
3
4
. Furfural → FA
. FA → THFA
. FA → MF
−35.95
−80.97
−25.27
104.78
0.19
0.21
0.13
−0.27
−0.83
200.38
391.69
211.06
577.44
396.81
−1.04
−1.01
−1.09
−1.50
−2.06
. MF → MTHF
Fig. 1 Reaction scheme for the hydrogenation of furfural.
5. THFA → MTHF 160.48
Green Chem.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Green Chemistry
Paper
the reduction of furfural to furfuryl alcohol over a copper
3
6
modified graphite felt cathode. Oxidation took place in a 1 M
NaCl + NaOH + LiOH anolyte and reduction in a phosphate
3 4
buffer (H PO + NaOH, pH 6.6) catholyte. In the half-cell
reactor, the electrocatalytic hydrogenation of furfural was
studied on Au, Pt, Ni, Cu, and Pb electrodes in different acidic
3
7
and basic media. It was determined that Cu was active
towards furfural conversion and that the highest conversion
was achieved in a basic solution.
From an economic viewpoint it would be desirable to do
the electrocatalytic hydrogenation of furfural in a continuous
flow reactor using stable electrodes without the use of a liquid
electrolyte. Previous work has suggested that hydrogenation is
3
7
most favorable in a basic media; however the Nafion mem-
brane used in PEM fuel cells is acidic. The environment in the
membrane electrode assembly (MEA) therefore limits the types
of metal catalysts that can be used for hydrogenation without
degradation. In this paper we will study the electrocatalytic
Fig. 2 Continuous electrocatalytic membrane reactor schematic for the electro-
catalytic hydrogenation of furfural over Pd/C or Pt/C using protons derived from
the electrolysis of water over Pt–Ru/C.
hydrogenation of 5 wt% furfural and 5 wt% furfuryl alcohol in using cyclic voltammetry conducted in an electrochemical
water in a continuous electrocatalytic membrane reactor with half-cell reactor using 20 wt% carbon supported metal at a
an MEA consisting of an acidic Nafion polymer electrolyte, Pt– loading of 1.0 mg cm− . The CVs were recorded within the
2
Ru/C anode, and Pd/C or Pt/C cathodes.
potential range of −0.2 to 1.0 V (vs. Ag/AgCl) at a scan rate of
0 mV s−
1
.
5
Continuous electrocatalytic membrane reactor experiments
were conducted using a PEM fuel cell (http://www.fuelcell-
technologies.com) with a 5 cm surface area. Serpentine flow
Experimental
2
The following chemicals were used in this study: furfural
channels for the anode and cathode were set in graphite
blocks, and liquid flow rates were controlled by Eldex® Optos
Metering Pumps. Water, at a flow of 0.6 ml min− , or 60 sccm
humidified hydrogen gas were used at the anode of the con-
tinuous electrocatalytic membrane reactor. The cathode con-
tained either 5 wt% furfural, 5 wt% furfuryl alcohol, or 5 wt%
furfural sparged with 30 sccm hydrogen gas. The liquid flow
rate at the cathode was 0.2 ml min− . The cathode residence
time at this flow rate was approximately 20 minutes. The elec-
trocatalytic reactor was purged with the anode and cathode
feeds for 20 minutes prior to applying a voltage to ensure the
cell was fully saturated with fresh feedstock. Samples were col-
lected from the cathode after an additional 20 minutes of flow
or once the current had reached steady-state. Gas samples at
(
C H O , 99% Acros Organics), furfuryl alcohol (C H O , 98%
5
4
2
5
6 2
1
Acros Organics), tetrahydrofurfuryl alcohol (C
Acros Organics), 2-methylfuran (C
and 2-methyltetrahydrofuran (C H O, 99% Acros Organics).
5
H
10 2
O , 99%
5
6
H O, 99% Acros Organics),
5
10
Liquid samples were analyzed by an Agilent 7890A gas chrom-
atography system, equipped with an RTX-VMS capillary
1
column and a flame ionization detector (FID). Helium was
used as the carrier gas with a column flow rate of 30 ml min−
1
.
One μl liquid sample was injected for each analysis. Carbon
balances were closed with 89 ± 2.5% for all experiments
reported here.
Electrocatalytic experiments
The schematic for the continuous electrocatalytic membrane the cathode outlet were collected by sparging the liquid
reactor system, including the feedstocks and products, is product with N . Voltage application and current measure-
shown in Fig. 2. Membrane electrode assemblies (MEAs) were ments were performed with a Metrohm Autolab Potentiostat,
2
−
2
fabricated using 4.0 mg cm Pt–Ru/C anodes and 1.0 mg PGSTAT302N, coupled with a 10 Amp Current Booster,
−
2
cm Pd/C or Pt/C cathodes. The working electrode used for BSTR10A. Gas samples were analyzed by a GC-5890 Hewlett
−
2
half-cell experiments was 1.0 mg cm Pd/C. Catalyst ink solu- Packard gas chromatography system equipped with a packed
tions consisted of 40 wt% Pt–Ru/C, 20 wt% Pd/C, or 20 wt% column and a thermal conductivity detector (TCD). Nitrogen
Pt/C (ETEK) mixed with deionized water, liquid Nafion was used as the carrier gas with a column pressure of 47 psi.
ionomer (Ion Power, Inc. Dupont DE520 Nafion ® Solution),
Electrocatalytic half-cell experiments, for the purpose of
and 1-propanol (99% Fisher). The ink solution was deposited determining active surface area of the catalyst, were conducted
on teflonized carbon paper (Toray, TGPH-090) using a spray using a standard three-electrode set-up. A platinum wire and
technique. MEAs were made using a hot-press at 140 °C and Ag/AgCl electrode served respectively as the counter and refer-
1
500 psi for 3 minutes. The membrane was Nafion 115 from ence electrodes. The catalyst ink solution for the working elec-
Nara Cell-Tech. The electrochemically active surface area of the trode was made in the same manner as described above. The
2
−1
2
−1
cathodes was measured as 44.7 m
g
and 48.2 m
g
2 4
for reaction solution was 5 wt% furfural in 0.5 M H SO . The half-
Pd/C and Pt/C, respectively. These measurements were taken cell was capped to maintain an oxygen-free environment, and
This journal is © The Royal Society of Chemistry 2013
Green Chem.

View Article Online
Paper
Green Chemistry
Fig. 3 Comparison of initial TOFs over different monometallic catalysts for (a) the APH of furfural at 100 °C and 800 psia for 75 minutes with 4.8 wt% furfural solu-
tion as feedstock, and (b) the APH of furfuryl alcohol at 80 °C and 800 psia for 3 hours with 4.8 wt% FA solution as feedstock.
2
the solution was purged with N for 10 minutes prior to intro- solutions were loaded into glass tube reactors containing
ducing H at a flow rate of 20 sccm. Gases were flowed into the reduced catalysts. The HTR was then sealed with closed valves
2
half-cell through a glass frit and the solution was stirred and transported to the reactor system bench. The HTR was
throughout the reaction to ensure adequate mixing.
first pressurized to 650 psia and then heated to 100 °C for the
APH of furfural and 80 °C for the APH of FA at a rate of 20 °C
−
1
High-throughput reactor (HTR) experiments
min , after which it was pressurized to a final pressure of
00 psia. The temperature and pressure were kept constant
8
Monometallic catalysts were prepared by the incipient wetness
impregnation method with the following metal precursors:
during the reaction, and a stirring speed of 800 rpm was used.
Once a reaction set finished, the HTR was cooled down to
room temperature at a rate of 14 °C min− . The HTR was then
depressurized to atmospheric pressure, and samples from
each well are taken and filtered at 0.2 microns.
[
Pd(NH
Ni(NO ) ·6H O (Strem Chemicals Inc.). The 5 wt% Rh/Al O
2 3
3
)
4
](NO
3
)
2
,
[Pt(NH
3
)
4
](NO
3
)
2
,
Ru(NO)(NO
3
)
3
and
1
3
2
2
catalyst was purchased from Strem Chemicals Inc. Gamma-
alumina (γ-Al ) was employed as a support, formed from
Boehmite (Sasol) pretreated in a conventional oven at 600 °C
for 4 hours. The catalysts were reduced under H flow at
00 °C for Pd, Pt, Ru, and Rh and 400 °C for Ni for 2 hours
2 3
O
2
3
Results
Catalyst selection
and purged with He flow for 30 minutes. After reduction, the
catalysts were cooled down to room temperature under He
flow, then sealed, and stored in a glove box. The metal disper- The initial activity of different alumina supported monometal-
sions of the prepared catalysts were determined by H chemi- lic catalysts including Pd, Pt, Ru, Rh, and Ni for the APH of
2
sorption using a Quantachrome Autosorb iQ Automated Gas furfural and FA was investigated. Fig. 3 shows the comparison
Sorption system. Before chemisorption, all catalysts were of initial turnover frequencies (TOFs) for the APH of furfural
reduced in situ under H flow up to 300 °C (Pd, Pt, Ru and Rh) and FA according to catalyst type. The initial activity for the
2
−
1
and 400 °C (Ni) with a heating rate of rate 1 °C min , held for APH of furfural decreased as follows: Pd > Ni > Ru > Pt > Rh (at
2
1
hours, purged with He for
40 minutes, and cooled to room temperature.
The aqueous-phase hydrogenation (APH) of furfural and FA APH of furfural. The order of initial activity for the APH of FA
was conducted using a HTR (HEL Group CAT24) consisting of followed: Pd > Ni > Ru > Rh ∼ Pt (at 80 °C). The 1 wt% Pd cata-
4 wells machined into a high-pressure stainless steel cylindri- lyst exhibited the highest initial TOF, which was 15.3 times as
2
hours, evacuated for 100 °C). The initial TOF of the 3 wt% Pd catalyst was 7.5 times
higher than the initial TOF of the 3 wt% Pt catalyst for the
2
cal chamber. The temperature, pressure, and stirring speed high as the initial TOF of the 1 wt% Pt catalyst. Ni also showed
were controlled and monitored by WinIso E670 system soft- a high initial activity for the APH of furfural and FA. However,
ware. Prior to the reaction, 5–10 mg of catalyst was loaded into Ni catalysts have been reported to solubilize in acidic environ-
a glass tube reactor placed on the HTR wells in the glove box ments and thus Ni is not a suitable monometallic PEM cata-
3
8
and the catalysts were reduced under H
2
flow at 200 °C for lyst. Although Pt had a low activity for furfural and FA
1
2 hours in case re-oxidation occurs. After reduction, the cata- hydrogenation, it was investigated further using the continu-
lysts were cooled down to room temperature under He flow, ous electrocatalytic membrane reactor due to its popularity as
then sealed, and moved to the glove box. In the glove box, both a fuel cell catalyst and as a conventional catalyst for fur-
3
2,33,39,40
magnetic stir bars and 2 ml of 4.8 wt% furfural and FA fural hydrogenation.
Green Chem.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Green Chemistry
Paper
Fig. 4 Current density and power input as a function of applied voltage for the electrocatalytic hydrogenation of furfural to form furfuryl alcohol and tetrahydro-
−1
−2
2
furfuryl alcohol over (a) Pd/C and (b) Pt/C. Reaction conditions: 30 °C, cathode: 111.7 h WHSV furfural over 1.0 mg cm Pd/C or Pt/C, anode: H O at 0.6 ml
−1
−2
min over 4.0 mg cm Pt–Ru/C.
Fig. 5 (a) Rate of furfural conversion and (b) current efficiency as a function of power input for the electrocatalytic hydrogenation of furfural. Reaction conditions:
0 °C, anode: H
2
O at 0.6 ml min− over 4.0 mg cm Pt–Ru/C, cathode: 111.7 h WHSV furfural over 1.0 mg cm Pd/C or 1.0 mg cm Pt/C.
1
−2
−1
−2
−2
3
The electrocatalytic hydrogenation of aqueous solutions of that does not contribute to furfural hydrogenation instead pro-
wt% furfural in a continuous electrocatalytic membrane duces hydrogen gas (that is not used to hydrogenate furfural)
5
reactor was investigated using both Pd/C and Pt/C cathodes, as at the cathode. The high current efficiency for the Pd/C catalyst
shown in Fig. 4 and 5. Fig. 4 shows the relationship between indicates that Pd/C is a better catalyst than Pt/C for furfural
current density and power input as a function of applied conversion, while Pt/C is a better catalyst for producing hydro-
voltage for the (a) Pd/C and (b) Pt/C cathode catalysts. The gen gas. This is further demonstrated in Fig. 5b, which depicts
power input is defined as the product of the applied voltage the rate of furfural conversion as a function of power input.
and the measured system current. The weight hourly space The rate of furfural conversion is up to 5 times higher with the
velocity (WHSV), defined as the ratio of the mass flow of fur- Pd/C catalyst as compared to the Pt/C catalyst. Additionally,
fural to the mass of the metal catalyst present in the MEA, was furfural hydrogenation over Pd/C at a power input of 0.05 W or
held constant for both of the catalytic systems. Both the more produced FA, THFA, MF, and MTHF. However, no MF or
current density and the power input are slightly higher for Pt/ MTHF was detected from furfural hydrogenation over Pt/C
C than for Pd/C. This shows that the Pt/C catalyst has a higher (results not shown). THFA was only present at a power input of
electrochemically active surface area and lower overpotential 0.3 W. Based on the higher rate of furfural conversion and
4
1
than Pd/C. The higher current density produced by Pt/C higher current efficiency, Pd/C was used as the cathode catalyst
translates directly to a higher flux of protons, using Faraday’s for all additional experiments.
constant, as compared to Pd/C.
Effect of applied voltage
The current efficiency at a given power input is up to 50
times higher for Pd/C compared to Pt/C as shown in Fig. 5a. The effect of applied voltage on the electrocatalytic hydrogen-
The current efficiency is the ratio of current contributing to ation of furfural using a Pd/C cathode was investigated over a
furfural hydrogenation divided by the overall current. Current voltage range of 1.15 V to 1.75 V. The results for current
This journal is © The Royal Society of Chemistry 2013
Green Chem.

View Article Online
Paper
Green Chemistry
Fig. 6 Selectivity as a function of (a) conversion and (b) applied voltage for the electrocatalytic hydrogenation of furfural over Pd/C. Reaction conditions: 30 °C,
−1
−2
−1
−2
2
cathode: 111.7 h WHSV furfural over 1.0 mg cm Pd/C, anode: H O at 0.6 ml min over 4.0 mg cm Pt–Ru/C.
efficiency (Fig. 5a) and reaction rate (Fig. 5b) are presented as
a function of total power input. The current efficiency ranged
from 24–30%. This signifies that 70–76% of the current con-
tributed to the evolution of hydrogen gas, rather than the
hydrogenation of furfural. Liquid product selectivity is
reported as a function of conversion in Fig. 6a and as a func-
tion of applied voltage in Fig. 6b. FA was the main product
with 100% selectivity at low conversion and applied voltage.
This result was expected given the theoretical standard cell
potential values in Table 1. As furfural conversion increased,
the FA selectivity decreased and other products were observed.
As shown in Table 1, a cell potential of 1.04 V is required for
converting furfural to FA, which is below the applied voltages
of 1.15 and 1.3 V. Converting FA to THFA or MF requires 1.01
and 1.09 V respectively. However, at low conversion there is
only low quantities of FA present and these products are not
observed. As conversion increases to 2%, THFA, MF, and
MTHF are detected at selectivities of 8–10%. Increasing the
Fig. 7 Reaction rate as a function of power input. Reaction conditions: 30 °C,
cathode: 111.7 h− WHSV furfural or 60 sccm N
1
over 1.0 mg cm Pd/C,
O at 0.6 ml min− over 4.0 mg cm Pt–Ru/C.
2
−2
2
1
−2
anode: H
applied voltage to 1.75 V increases conversion to 6% under the rate of hydrogen production for the H
reaction conditions used in this study. At this voltage, THFA decreased compared to the rate of hydrogen production for the
was the second most abundant product with a selectivity of
O–N system. The rate of furfural conversion in the H O–fur-
6%. MF and MTHF remain at 8% selectivity. The high selecti- fural system increased with increasing power input, but is up
2
O–furfural system
H
2
2
2
2
vity for THFA, as compared to MF or MTHF, was expected to 16 times lower than the rate of hydrogen production in the
based on the thermodynamic calculations. The high standard same system. Also shown in Fig. 7 is the rate of hydrogen
cell potential (2.06 V) for THFA hydrogenation to MTHF is con- production plus the rate of furfural conversion for the H
sistent with these experimental results and indicates that the furfural system. This “combined rate” is similar to the rate of
main route for formation of MTHF is through MF.
hydrogen production in the O–N system, especially
2
O–
H
2
2
Fig. 7 compares the rate of hydrogen production, furfural for power inputs below 0.1 W. Thus, it can be inferred that,
conversion, and hydrogen production plus furfural conversion at power inputs below 0.1 W, the furfural does not inhibit
(
“combined rate”) as a function of power input for two the rate of water electrolysis. The difference between
different systems: (1) H O on the anode/furfural on the hydrogen production in the H O–N system and the “com-
O–furfural system increases with
system). Hydrogen gas increasing power input. This difference coincides with a
was collected at the cathode by sparging the liquid product decrease in the current density of the H O–furfural system.
with nitrogen gas. The highest measured rate of hydrogen pro- One possible cause could be that the liquid furfural feedstock
duction was from the H O–N
system. At power inputs less at the cathode of the H O–furfural system results in a higher
than 0.05 W, the rate of hydrogen production was comparable over potential, and thus a lower current, than the gaseous N
2
2
2
cathode (H
hydrated N
2
O–furfural system); and (2) H
on the cathode (H O–N
O on the anode/ bined rate” in the H
2
2
2
2
2
2
2
2
2
2
for both systems. However, as the power input increases, the in the H
O–N system.
2
2
Green Chem.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Green Chemistry
Paper
Fig. 8 (a) Current density, current efficiency and (b) furfural conversion rate, power input as a function of temperature for the electrocatalytic hydrogenation of fur-
fural. Reaction conditions: 1.45 V, cathode: 111.7 h− WHSV furfural over 1.0 mg cm Pd/C, anode: H
1
−2
O at 0.6 ml min over 4.0 mg cm Pt–Ru/C.
−1
−2
2
Fig. 9 (a) Reactant concentration as a function of applied voltage and (b) reactant conversion rate as a function of power input for the electrocatalytic hydrogen-
−1
ation of furfuryl alcohol. Furfural conversion is included in Fig. 9b for the purpose of comparison. Reaction conditions: 30 °C, cathode: 111.7 h WHSV furfuryl
−
1
−2
−1
−2
alcohol (or 111.7 h WHSV furfural) over 1.0 mg cm Pd/C, anode: H
2
O at 0.6 ml min over 4.0 mg cm Pt–Ru/C.
Effect of reactor temperature
cell to further increase temperature and study its effects on
reaction rate and efficiency.
4
2,43
The temperature of the continuous electrocatalytic membrane
reactor was increased from 30 °C to 70 °C to investigate the
effect of reaction temperature. Fig. 8a shows that the current
Electrocatalytic hydrogenation of furfuryl alcohol
density increases while the current efficiency decreases, with Fig. 9 depicts results from the electrocatalytic hydrogenation
increasing temperature. The decreasing current efficiency sig- of FA. Fig. 6 shows that at low conversion and low applied
nifies an increase in the production of hydrogen gas. These voltage only FA was produced for furfural conversion. It was
results indicate that the increase in temperature promotes explained that sufficient voltage was applied for producing MF
water electrolysis and production of hydrogen gas but not fur- and THFA but that insufficient FA was present for these reac-
fural hydrogenation. This is confirmed by Fig. 8b, which tions. At higher voltages, where there was a higher degree of
shows no change in the furfural conversion rate, despite the furfural conversion, MF and THFA were produced. Given this
increase in power input. There is also no statistically signifi- explanation, the electrocatalytic reduction of FA should
cant change in liquid product selectivity as the temperature produce MF and THFA at an applied voltage of 1.15 V. This is
increases (results not shown). In non-electrocatalytic systems confirmed by the results in Fig. 9a, which show that MF and
furfural hydrogenation is generally conducted at temperatures THFA are produced at approximately 3.8 mM at an applied
4
1
greater than 120 °C. The Nafion membrane used in the con- voltage of 1.15 V. However, at 1.15 V no MTHF is observed,
tinuous electrocatalytic membrane reactor is only stable up to likely because there is insufficient MF present for further reac-
8
0 °C, and therefore, additional experiments could be con- tion. FA conversion and concentrations of THFA and MTHF
ducted using a high temperature membrane, such as a increase as applied voltage increases. However, the concen-
Celtec®P membrane from BASF, or a high temperature fuel tration of MF increases from 1.15 V to 1.45 V, but then
This journal is © The Royal Society of Chemistry 2013
Green Chem.

View Article Online
Paper
Green Chemistry
decreases when voltage is increased to 1.75 V. This trend
appears to further confirm that MF is an intermetidate for
MTHF, as was discussed for Fig. 6b.
Discussion
The main challenge in electrocatalytic hydrogenation using a
The FA conversion rate increased with increasing power continuous electrocatalytic membrane reactor is that a large
input, as shown in Fig. 9b. Initially, the rate of FA conversion fraction of the current goes towards generating hydrogen gas
is higher than the rate of furfural conversion. Furfural hydro- that is not used in the hydrogenation reaction. Thus, to make
genation produces primarily FA at low power input, which this system more practical is highly desirable to increase the
involves only hydrogenating the aldehyde functionality. FA rate of hydrogen consumption due to hydrogenation reactions
hydrogenation produces THFA, which involves hydrogenating over the rate of hydrogen gas evolution. This is based not only
the furan ring, and MF, which involves hydrogenolysis of the on the low rate of furfural hydrogenation (0.6 moles furfural
hydroxyl group. Therefore, the rate of hydrogenating the alde- per moles Pd min), but also the low current efficiency
hyde functionality in furfural appears lower than the sum of (24–30%), which indicates that the majority of protons are
the rates of hydrogenating the furan ring and hydrogenolysis going towards hydrogen gas evolution. One method of increas-
of the hydroxyl group in FA. At power inputs greater than 0.12 ing the furfural hydrogenation rate is to increase the applied
W, the rate of furfural conversion is higher than the rate of FA voltage, which equates to an increase in power input. However,
conversion. This result indicates that the rate of furfural con- this also increases the rate of hydrogen evolution and thereby
version increases due to the added reactions of FA to THFA, FA affects the efficiency of the reactor. It was shown in Fig. 6b that
to MF, and MF to MTHF.
current efficiency for furfural conversion is between 24–30%,
and increases with increasing power input. Power efficiency is
defined as the sum of the higher heating value of the hydro-
genated products divided by the sum of the higher heating
value of the reactants plus the electrical power input. Fig. 11
shows that there is an initial drop in power efficiency as a func-
tion of power input, followed by a graduate decrease at inputs
greater than 0.05 W. Therefore, a small decrease in current or
power efficiency may be a worthwhile penalty for a higher rate
of furfural hydrogenation. However, the applied voltage also
affects selectivity, and thus increasing it may increase overall
reaction rate but not promote the desired product (ex. FA at
Electrocatalytic hydrogenation using hydrogen gas
The rate of furfural hydrogenation at low power input using
protons derived from hydrogen gas (the H –furfural system)
was found to be comparable to the rate of furfural hydrogen-
ation at higher power input using protons derived from water
2
electrolysis (the H O–furfural system), as shown in Fig. 10.
2
This is similar to results obtained for the electrocatalytic
hydrogenation of acetone. Fig. 10 also shows that the H –fur-
2
fural system has a lower current efficiency, at all investigated
power inputs, as compared to the H O–furfural system. There-
1
0
2
1.3 V vs. FA at 1.75 V as seen in Fig. 6). One method of manipu-
2
fore, the H –furfural system not only has a higher rate of fur-
lating applied voltage without influencing selectivity is to
change the species being oxidized at the anode. For example,
Rodriguez-Rivera et al. found that the rate of cathodic hydro-
gen gas evolution is comparable for oxidation of an aqueous
solution of polyoxometalates (POM), which occurs at ∼0.62 V,
fural conversion but also a higher rate of hydrogen gas
production. Both systems have comparable liquid product
selectivities of MF and MTHF. Selectivity of THFA is slightly
higher in the H –furfural system and selectivity of FA is slightly
2
2 2
lower in the H –furfural system, as compared to the H O–fur-
4
4
as compared to oxidation of water at ∼2.0 V.
fural system (see ESI†).
Fig. 10 Current efficiency (CE) and furfural conversion rate as a function of
power input for the electrocatalytic hydrogenation of furfural using hydrogen
Fig. 11 Power efficiency as a function of power input for the electrocatalytic
−1
−1
gas or water at the anode. Reaction conditions: 30 °C, cathode: 111.7 h
hydrogenation of furfural. Reaction conditions: 30 °C, cathode: 111.7 h WHSV
−2
−1
−2
−1
−2
WHSV furfural over 1.0 mg cm Pd/C, anode: H
2
O at 0.6 ml min or 30 sccm
2
furfural over 1.0 mg cm Pd/C, anode: H O at 0.6 ml min over 4.0 mg cm
−
2
H
2
over 4.0 mg cm Pt–Ru/C.
Pt–Ru/C.
Green Chem.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Green Chemistry
Paper
Fig. 12 Reactor schematics for the hydrogenation of furfural using protons generated from water electrolysis. (a) PEM water electrolyzer coupled with a conven-
tional catalytic reactor. Hydrogen gas is produced from water electrolysis in the PEM electrolyzer, which is then used in a conventional catalytic reactor to hydrogen-
ate furfural. (b) Single continuous electrocatalytic membrane reactor with second catalytic bed (electrocatalytic or conventional catalytic) and recycle loops.
A second method of increasing the rate of furfural hydro- to recompress the hydrogen gas after reaction. In addition,
genation is to increase the reactor temperature. However, the this approach requires a separate high-pressure reactor. The
temperature of the continuous electrocatalytic membrane continuous electrocatalytic membrane reactor proposed in this
reactor is limited by the Nafion membrane and, as discussed study was depicted in Fig. 2. In this design, water is electro-
earlier, either a high temperature membrane or another type lyzed at the anode of the electrocatalytic reactor and the gener-
4
2,43
of fuel cell is required to achieve temperatures over 80 °C.
ated protons are used to reduce furfural at the cathode. The
Additionally, increasing reactor temperature was seen to benefit of this design, as compared to that shown in Fig. 12a,
decrease current efficiency (Fig. 8a). A third option to increase is that it does not require pressurization, storage, or transport
reaction rate is to change the catalyst or its support. Ni, Fe, of hydrogen gas and takes place all in one reactor, thereby
and Cu have been used for electrocatalytic furfural hydrogen- minimizing energy and materials usage.
ation and shown high conversion and selectivity to FA, despite
The continuous electrocatalytic membrane reactor could be
3
4
leaching of these metals. These metals could be readily further developed to recycle hydrogen and/or furfural from the
incorporated into an MEA for an alkaline fuel cell. Using con- cathodic product stream and/or to include a second catalytic
ventional catalysis, it has been shown that Pt/titania results in bed to further hydrogenate remaining furfural (Fig. 12b). The
higher selectivity to FA and THFA while Pt/silica and Pt/magne- second catalytic bed could be conventional, and thus require a
sia have higher selectivity towards MF, 2-pentanol, and 2- compressor, or electrocatalytic. A recycle loop for water exiting
3
2
pentanone.
Extensive research has been conducted on the anode could also be incorporated. The purpose of recycling
improving the kinetics for H /O fuel cells, which could be hydrogen gas in the cathode feedstock is not only to conserve
2
2
applicable to electrocatalysis of biomass derived molecules. resources but also to increase the rate of furfural hydrogen-
Research to improve oxygen reduction reactions (ORR) has ation. However, experiments conducted with hydrogen gas
examined Pt monolayers on Au, Rh, Pd, Ru, and Ir supports as sparged at the cathode inlet showed no statistically relevant
well as monolayers of Pt–M (M = Ir, Ru, Rh, Pd, Au, Re, or Os) changes in furfural conversion rate or product selectivity (see
deposited on the surfaces of Pd (111) single crystal or carbon ESI†).
4
5,46
supported Pd nanoparticles.
To decrease CO poisoning
and improve reaction kinetics at the anode, Pt alloys such as
4
7,48
PtRu, PtMo, and PtSn have been examined.
An added
Conclusion
difficulty in choosing the proper catalyst is that the catalyst
represents
process.
a
large portion of the cost of the overall In this paper we studied the electrocatalytic hydrogenation of
The aforementioned Pt monolayers drastically furfural in a continuous electrocatalytic membrane reactor
4
9,50
decrease the amount, and therefore, cost of Pt in the catalyst. using protons derived from water electrolysis. Pd was 7.5 times
Even cheaper substrates made from tungsten carbide, have more active than Pt for hydrogenation of furfural in a batch
also been found to have a similar hydrogen binding energy reactor. Consistent with the batch reactor experiments, Pd/C
and activity to bulk Pt, when only a Pt monolayer is was 4.4 times more active than Pt/C for hydrogenation of fur-
5
1,52
present.
fural at the cathode of the continuous electrocatalytic mem-
Alternate reactor designs could also be used for this brane reactor. Identified products from furfural hydrogenation
process, as shown in Fig. 12. One of the most basic methods over Pd/C were FA, THFA, MF, and MTHF. The product selecti-
of hydrogenating furfural would be to produce hydrogen in a vity changed as a function of applied voltage. The FA selectivity
PEM electrolyzer and then use the hydrogen in a conventional decreased and the THFA selectivity increased with increasing
catalytic reactor, as seen in Fig. 12a. The disadvantage of this voltage. Selectivity towards MF and MTHF was 8% at power
design is that the hydrogenation of furfural requires high- inputs over 0.02 W. The current efficiency was 24–30%, which
pressure hydrogen. Therefore a compressor would be needed indicates that up to 76% of the overall current contributed
This journal is © The Royal Society of Chemistry 2013
Green Chem.

View Article Online
Paper
Green Chemistry
towards the evolution of hydrogen gas instead of furfural 17 R. P. O’Hayre, S. W. Cha, W. Colella and F. B. Prinz, Fuel Cell
hydrogenation. Adjusting the temperature between 30–70 °C Fundamentals, John Wiley & Sons, New York, 2006, ch. 3.
was not found to affect the rate of furfural conversion. The 18 A. J. Bard and L. F. R, Electrochemical Methods: Fundamen-
electrocatalytic hydrogenation of FA resulted in high selecti-
vities of THFA and MF, and a faster initial rate of conversion
tals and Applications, John Wiley & Sons, New York, 2nd
edn, 2001, ch. 3.
than furfural hydrogenation. Using hydrogen gas as the source 19 Z. Zhang, L. Xin and W. Li, Appl. Catal., B, 2012, 119–120,
of protons instead of water electrolysis resulted in a compar- 40–48.
able rate of furfural conversion at a lower power input. Spar- 20 L. Xin, Z. Zhang, Z. Wang and W. Li, ChemCatChem, 2012,
ging hydrogen gas at the cathode had no statistically
4, 1105–1114.
significant effect on the reaction rate.
21 A. Corma, S. Iborra and A. Velty, Chem. Rev., 2007, 107,
2411–2502.
2
2
2 S. F. Paul, Alternative Fuels, 2001.
3 R. Xing, W. Qi and G. W. Huber, Energy Environ. Sci., 2011,
Acknowledgements
4, 2193.
This work is supported as part of the Catalysis Center for 24 T. Werpy and G. Petersen, Top Value Added Chemicals from
Energy Innovation, an Energy Frontier Research Center funded
by the U.S. Department of Energy, Office of Science, Office of
Basic Energy Sciences under Award Number DE-SC0001004
and by the University of Massachusetts CVIP Technology
Biomass Volume I—Results of Screening for Potential Candi-
dates from Sugars and Synthesis Gas Top Value Added Chemi-
cals From Biomass Volume I: Results of Screening for Potential
Candidates, 2004.
Development fund. Catalyst supplies were obtained by the 25 M. a. Tike and V. V. Mahajani, Ind. Eng. Chem. Res., 2007,
support of the Ministry of Knowledge Economy (MKE) (2010- 46, 3275–3282.
H-003-00030100-2010) in South Korea. The authors would also 26 R. Rao, A. Dandekar, R. T. K. Baker and M. A. Vannice,
like to acknowledge Professor Datta from Worcester Polytech-
nic Institute for use of his equipment in fabricating the MEAs.
J. Catal., 1997, 171, 406–419.
27 R. S. Rao, R. T. K. Baker and M. A. Vannice, Catal. Lett.,
1999, 60, 51–57.
2
2
8 J. Lessard, J.-F. Morin, J.-F. Wehrung, D. Magnin and
E. Chornet, Top. Catal., 2010, 53, 1231–1234.
9 L. Baijun, L. Lianhai, W. Bingchun, C. Tianxi and
K. Iwatani, Appl. Catal., A, 1998, 171.
References
1
2
3
4
5
G. W. Huber, S. Iborra and A. Corma, Chem. Rev., 2006,
06, 4044–4098.
J. N. Chheda, G. W. Huber and J. A. Dumesic, Angew.
Chem., Int. Ed., 2007, 46, 7164–7183.
1
30 A. B. Merlo, V. Vetere, J. F. Ruggera and M. L. Casella,
Catal. Commun., 2009, 10, 1665–1669.
31 V. Vetere, A. B. Merlo, J. F. Ruggera and M. L. Casella,
J. Braz. Chem. Soc., 2010, 21, 914–920.
32 J. Kijeński, P. Winiarek, T. Paryjczak, A. Lewicki and
A. Mikołajska, Appl. Catal., A, 2002, 233, 171–182.
33 P. D. Vaidya and V. V. Mahajani, Ind. Eng. Chem. Res., 2003,
42, 3881–3885.
34 Z. Li, S. Kelkar, C. H. Lam, K. Luczek, J. E. Jackson, D. J. Miller
and C. M. Saffron, Electrochim. Acta, 2012, 64, 87–93.
D. M. Alonso, J. Q. Bond and J. A. Dumesic, Green Chem.,
2
010, 12, 1493.
J. C. Serrano-Ruiz and J. A. Dumesic, Energy Environ. Sci.,
011, 4, 83–99.
N. Li, G. A. Tompsett and G. W. Huber, ChemSusChem,
010, 3, 1154–1157.
D. C. Elliott, Energy Fuels, 2007, 21, 1792–1815.
2
2
6
7
T. P. Vispute and G. W. Huber, Green Chem., 2009, 11, 35 D. Chu, Y. Hou, J. He, M. Xu, Y. Wang, S. Wang, J. Wang
433–1445. and L. Zha, J. Nanoparticle Res., 2009, 11, 1805–1809.
J. Monnier, H. Sulimma, A. Dalai and G. Caravaggio, Appl. 36 G. Chamoulaud, D. Floner, C. Moinet, C. Lamy and
Catal., A, 2010, 382, 176–180. E. M. Belgsir, Electrochim. Acta, 2001, 46, 2757–2760.
C. E. G. Padró and V. Putsche, Survey of the Economics of 37 P. Parpot, A. P. Bettencourt, G. Chamoulaud, K. B. Kokoh
Hydrogen Technologies, Golden, Colorado, 1999. and E. M. Belgsir, Electrochim. Acta, 2004, 49, 397–403.
0 S. K. Green, G. A. Tompsett, H. J. Kim, W. B. Kim and 38 Y. Shao, G. Yin and Y. Gao, J. Power Sources, 2007, 171, 558–
G. W. Huber, ChemSusChem, 2012, 5, 2410–2420. 566.
1 P. Choi, D. G. Bessarabov and R. Datta, Solid State Ionics, 39 P. Costamagna and S. Srinivasan, J. Power Sources, 2001,
1
8
9
1
1
2
004, 175, 535–539.
102, 242–252.
1
1
2 W. Smith, J. Power Sources, 2000, 86, 74–83.
3 H. Chen, T. N. Cong, W. Yang, C. Tan, Y. Li and Y. Ding,
Prog. Nat. Sci., 2009, 19, 291–312.
40 T. B. L. W. Marinelli, V. Ponec, C. G. Raab and J. A. Lercher,
Heterogeneous Catalysts and Fine Chemicals III, 1993,
pp. 195–202.
1
1
1
4 F. Barbir, Sol. Energy, 2005, 78, 661–669.
5 N. Li and G. W. Huber, J. Catal., 2010, 270, 48–59.
6 N. Li, G. A. Tompsett, T. Zhang, J. Shi, C. E. Wyman and 42 R. J. Gorte and J. M. Vohs, Ann. Rev. Chem. Biomol. Eng.,
G. W. Huber, Green Chem., 2011, 13, 91. 2011, 2, 9–30.
41 S. a. Grigoriev, P. Millet and V. N. Fateev, J. Power Sources,
2008, 177, 281–285.
Green Chem.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Green Chemistry
Paper
4
3 A. Jayakumar, R. Küngas, S. Roy, A. Javadekar, D. J. Buttrey, 48 J. Greeley, J. K. Nørskov, L. a. Kibler, A. M. El-Aziz and
J. M. Vohs and R. J. Gorte, Energy Environ. Sci., 2011, 4,
133.
4 G. J. Rodriguez-Rivera, W. B. Kim, S. T. Evans, T. Voitl and
D. M. Kolb, ChemPhysChem, 2006, 7, 1032–1035.
49 S. K. Kamarudin, W. R. W. Daud, A. Md. Som, M. S. Takriff
and A. W. Mohammad, J. Power Sources, 2006, 157,
641–649.
4
4
4
J. A. Dumesic, J Am. Chem. Soc., 2005, 127, 10790–10791.
5 J. Zhang, M. B. Vukmirovic, K. Sasaki, A. U. Nilekar, 50 I. Bar-on, R. Kirchain and R. Roth, Science, 2004, 109,
M. Mavrikakis and R. R. Adzic, J. Am. Chem. Soc., 2005,
27, 12480–12481.
6 J. Zhang, M. B. Vukmirovic, Y. Xu, M. Mavrikakis and
71–75.
1
51 D. V. Esposito, S. T. Hunt, A. L. Stottlemyer, K. D. Dobson,
B. E. McCandless, R. W. Birkmire and J. G. Chen, Angew.
Chem., Int. Ed., 2010, 49, 9859–9862.
4
R. R. Adzic, Angew. Chem., Int. Ed., 2005, 44, 2132–2135.
4
7 A. U. Nilekar, K. Sasaki, C. A. Farberow, R. R. Adzic and 52 D. V. Esposito, S. T. Hunt, Y. C. Kimmel and J. G. Chen,
M. Mavrikakis, J. Am. Chem. Soc., 2011, 133, 18574–18576. J. Am. Chem. Soc., 2012, 134, 3025–3033.
This journal is © The Royal Society of Chemistry 2013
Green Chem.